Time-Domain Telescope

Sep 30, 2009 21:45

So, apparently, researchers have invented a method for transferring data at an increased rate using what they refer to as a "time-domain telescope", a way of focusing optical pulses to make them "shorter in time" to transmit data at an increased rate. Unfortunately, no article I have found seems to be able to use unambiguous language to describe ( Read more... )

rant, science!

Leave a comment

Comments 28

firynze October 1 2009, 02:28:29 UTC
all I'm finding on my usual photonics resource site is older articles, nothing after 2008...

Reply

smarriveurr October 1 2009, 02:36:56 UTC
... my usual photonics resource site...

This is why I love LJ. My friends are awesome.

Reply

firynze October 1 2009, 12:45:55 UTC
When in doubt, Laurin Photonics.

Reply

smarriveurr October 1 2009, 14:48:50 UTC
Duly noted. And they have a glossary!... albeit poorly implemented.

Reply


magdelane October 1 2009, 03:50:36 UTC
hm. Race=relay?
If I remember, I bet I can get that article at our library. it would be better if it is one that's available as an online journal. We'll see. Tomorrow is looking to be a crazy but fun day.

Reply

smarriveurr October 1 2009, 14:38:03 UTC
I managed to read the abstract for the original Nature Photonics article, but it doesn't even link to a "subscribe now to read this article" sort of thing like most sites would do. I assume that, if NP puts it online, it's paid subscription only - and herself already checked whether the college had an access subscription. ;)

Reply

magdelane October 1 2009, 15:10:23 UTC
Ah, well.
I've actually heard about this system, it sounds much like one of the options that is being considered for Constellation's network and data transmission, and something TDRS would take on for it. Seeing that's what we do here, we might have access to the journal. Just sayin'.

Reply

smarriveurr October 1 2009, 15:15:48 UTC
Again, this is why I love my friends.

It would make sense, since we're talking about multiplying data transmission rates by a fairly significant amount at the high end. And if you do find the article, I'm naturally dying for more info. I've got access to a photonics glossary now, so there's even a chance I could understand it!

I mean, it's not like it's rocket science, right? ;)

Reply


paperkingdoms October 1 2009, 04:05:13 UTC
Yeah, I think they're saying that the front part is slowed down, and the rear part is sped up, so that overall it takes less time. A race is a pretty goofy way of saying it. It's roughly analogous to saying that the image that you get if you shrunk a picture with a lens was done by having the edges of the image scoot over toward the middle.

I'm a little curious, because the language of some of it early on is a little bit applied-what-I-do-ish. But not nearly enough to be helpful, really. [Nor am I applied enough to be helpful. But you knew that.]

Reply

smarriveurr October 1 2009, 14:43:16 UTC
*nod* The language in pretty much all of these is not so much "simplified" as "obfuscatory". I can't help but feel they could benefit from differentiating "length in time" from "duration", "speed of transport" from "rate of data transmission", etc, etc. I mean, I assume the journalists aren't necessarily going to grok what's really happening, but you'd think there'd be some middle-ground.

And, well, if I find a better article, I'll let you know. Alternately, you can check whether your uni or library have a subscription to Nature Photonics, where the original research was published. And that would be awesome, because I've waited about two decades to be able to say "See what you could accomplish if you applied yourself?" to someone else.

Reply

paperkingdoms October 1 2009, 17:45:56 UTC
We don't have it at KU, but I put in a request. I should get a scan of it sometime soon, and I'll share.

Reply

smarriveurr October 1 2009, 17:52:45 UTC
I posted a little while back about it, but I think I've at least grasped the idea on the "lies-to-children" level, which is probably as much as I'm likely to accomplish without a few more years of academia.

Basically, they send the data pulse through the waveguide with an infrared pulse. The infrared pulse vibrates the atoms of the silicon waveguide, which alter the frequency of portions of the data pulse. Since the higher frequency light can travel faster through the medium, the researchers can compress the pulse by accelerating the rear end and decelerating the leading end. Beyond that, I don't know what the hell is happening, but that much at least makes more sense than before, and is kind of like what I thought was happening.

I will, however, totally take that scan to see if I can grok any more in the original. Thanks!

Reply


Leave a comment

Up